Figure. Microbial biofilms contain diverse bacteria, including many known uropathogens, in both infected and noninfected devices. Shown here are the genus-level profiles of biofilms from each explanted inflatable penile prosthesis device, separated by infection status. Bars relating to the most commonly detected uropathogens are indicated for each device.

Inflatable penile prostheses (IPP) are an effective treatment option for patients with erectile dysfunction. These devices consist of multiple components that include 2 cylinders, a pump, and a reservoir. However, there are variations of this configuration, which include malleable prosthetic devices. The number of components and complexity of devices make multiple niches available for the colonization and proliferation of bacterial biofilms that can cause infectious complications. Thus, despite the high degree of efficacy and patient satisfaction with IPPs, a small subset of devices still require removal due to infection. Importantly, infection is associated with significant morbidity and costs that exceed initial IPP placement by sixfold.¹ Given the considerable negative implications that IPP infections present, several strategies to prevent infection have been developed, including the use of antibiotic-coatings/impregnation, which has been associated with reduced infection risk.^1,2 However, despite the implementation of such strategies, as well as modifications of surgical technique and prophylaxis, there remains a 1%-3% risk of infection.^1,3-5

In this study, we set out to determine the presence and composition of biofilms adherent to IPP devices in both infected and noninfected devices to better understand the natural history of biofilms and the transition from the asymptomatic to infectious state. We hypothesized that IPP devices would harbor unique biofilms based on factors such as infection status, antibiotic use, or indwelling time.

A total of 27 patients scheduled for penile prosthesis removal or revision were identified and consented per the institutional review board–approved protocol. Inclusion criteria included individuals over 18 years of age undergoing removal or revision of an IPP device. Upon surgical removal, the first accessed portion of the device was swabbed with standard culture swabs for culture-based and molecular microbiome analyses. Swabbed specimens from each device were maintained at 4 °C and transferred to –80 °C within 4 hours of collection, prior to processing. Multiple positive and negative controls were obtained and analyzed in tandem.

The mean patient age at device explantation was 64 (±11.5) years, and mean device indwelling time was 4.8 (±5.0) years. A total of 41% of the population from whom devices were explanted had a diagnosis of diabetes mellitus, and 11% were current smokers. 4 patients (14.8%) had evidence of device-associated infection requiring explantation and 4 patients had device-associated pain requiring explantation (14.8%).

Molecular analysis found that 93% of penile prostheses samples had a statistically relevant bacterial load, indicative of an active biofilm on devices. While the number of unique bacterial species detected increased significantly with device indwelling time (P = .002), they did not differ as a function of age or operative time at device placement. This suggests that adherent biofilms may originate from the native microbiota rather than introduction through a sterility breach during surgical implantation.

Diagnoses of diabetes mellitus, cardiac disease, or current smoking status were not associated with significant differences in the number of unique species detected. However, we found that the microbiome composition differed between those who had an antibiotic prescription within 30 days of device explantation vs those who did not (P = .05). Importantly, some common uropathogens such as Klebsiella (log₂ fold change +4.4, P < .001) were more abundant in biofilms from individuals who had antibiotics within 30 days of device removal. These data support the model wherein antibiotic use may shift the composition of these biofilms to promote rather than eliminate the presence of uropathogenic bacteria. We recently reported a similar increase in abundance of uropathogens with antibiotic use for indwelling stents.⁶

Interestingly, composition did not differ based on device-associated infection or device-associated pain. While infected devices generally harbored known uropathogens that made up a large fraction of biofilms, nearly all devices with a detectable microbiome harbored at least 1 known uropathogen, sometimes comprising a significant proportion of the biofilm (see Figure). These data demonstrate that the mere presence of a known pathogen does not equate to an infection. Rather, in a subset of cases, infections may result from the environment promoting specific infectious processes.

While the current clinical paradigm to manage infectious complications for implantable devices aims to prevent all bacterial colonization on devices, the data from our study suggest that nearly all devices had a detectable biofilm with 1 or more known uropathogens present. Importantly, one of the most common management practices, the administration of antibiotics, actually increased rather than decreased the abundance of some uropathogens. To better manage infectious complications, our data suggest that we need to shift our clinical paradigm from the thought that any bacterial colonization equates to an infection, to one in which bacterial colonization is common and infections result in a subset of cases from specific microbial processes, given the right environment. Instead of using broad spectrum therapies intended to indiscriminately eliminate bacteria, the optimal targets may be specific infectious processes to prevent harmful complications while leaving the native biofilms intact.